xlayn's submissions | Hacker News

1.		Qwen3.8-27B make medium the default effort level instead of xhigh (github.com/alainnothere)
		16 points by xlayn 33 days ago \| past \| 4 comments
2.		Privibe – LLM CLI Local first+privacy focus+llama.cpp cache branch and Qwen3.x (github.com/alainnothere)
		1 point by xlayn 34 days ago \| past \| 1 comment
3.		Privibe - LLM Cli Local first+privacy focus+llama.cpp cache branch + Qwen3.x (github.com/alainnothere)
		2 points by xlayn 36 days ago \| past \| 1 comment
4.		Ask HN: What's slop? what's AI written text and why read/not read?
		7 points by xlayn 38 days ago \| past \| 7 comments
5.		Copilot "auto-pilot" system instructions making models worst
		2 points by xlayn 4 months ago \| past
6.		AMD GPU LLM Performance Testing (github.com/alainnothere)
		4 points by xlayn 5 months ago \| past \| 1 comment
7.		Show HN: Duplicate 3 layers in a 24B LLM, logical deduction .22→.76. No training (github.com/alainnothere)
		265 points by xlayn 6 months ago \| past \| 80 comments
8.		Tell HN: CryptoWorker Running on Telegram Web
		3 points by xlayn on Sept 21, 2023 \| past
9.		Ask HN: What should I learn to be competitive in the next five years?
		9 points by xlayn on Aug 1, 2015 \| past \| 7 comments